The fixed-point algorithm and maximum likelihood estimation for independent component analysis

被引：189

作者：

Hyvärinen, A ^{[1
]}

机构：

[1] Aalto Univ, Lab Comp & Informat Sci, FIN-02015 Espoo, Finland

来源：

NEURAL PROCESSING LETTERS | 1999年 / 10卷 / 01期

关键词：

independent component analysis; blind source separation; maximum likelihood;

D O I：

10.1023/A:1018647011077

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The author previously introduced a fast fixed-point algorithm for independent component analysis. The algorithm was derived from objective functions motivated by projection pursuit. In this paper, it is shown that the algorithm is closely connected to maximum likelihood estimation as well. The basic fixed-point algorithm maximizes the likelihood under the constraint of decorrelation, if the score function is used as the nonlinearity. Modifications of the algorithm maximize the likelihood without constraints.

引用

页码：1 / 5

页数：5

共 13 条

[1]

Amari S, 1997, ADV NEUR IN, V9, P127

[2] AN INFORMATION MAXIMIZATION APPROACH TO BLIND SEPARATION AND BLIND DECONVOLUTION [J].

BELL, AJ ;

SEJNOWSKI, TJ .

NEURAL COMPUTATION, 1995, 7 (06) :1129-1159

[3]

CARDOSO JF, 1998, ADAPTIVE UNSUPERVISE

[4] Robust neural networks with on-line learning for blind identification and blind separation of sources [J].

Cichocki, A ;

Unbehauen, R .

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 1996, 43 (11) :894-906

[5] INDEPENDENT COMPONENT ANALYSIS, A NEW CONCEPT [J].

COMON, P .

SIGNAL PROCESSING, 1994, 36 (03) :287-314

[6] PROJECTION PURSUIT [J].

HUBER, PJ .

ANNALS OF STATISTICS, 1985, 13 (02) :435-475

[7] A fast fixed-point algorithm for independent component analysis [J].

Hyvarinen, A ;

Oja, E .

NEURAL COMPUTATION, 1997, 9 (07) :1483-1492

[8]

HYVARINEN A, IN PRESS IEEE T NEUR

[9]

HYVRINEN A, 1997, P IEEE INT C AC SPEE, P3917

[10] BLIND SEPARATION OF SOURCES .1. AN ADAPTIVE ALGORITHM BASED ON NEUROMIMETIC ARCHITECTURE [J].

JUTTEN, C ;

HERAULT, J .

SIGNAL PROCESSING, 1991, 24 (01) :1-10

← 1 2 →